M. L. TiLDESLEY 
21Y 
measurements, has no doubt been made by all users of Ranke's craniophor. It 
will be necessary therefore to deduct 2-5 mms. from all direct measurements of 
auricular height given by Fawcett*, Macdonnell*, Benington| and Thomson J. 
Auricular Height. {First five corrected.) 
Race Male Female 
Naqada 113-1 110-6 
Whitechapel Englisli 112-1 106-7 
Liverpool St English 111-3 106-9 
Congo Negroes ... 111-4 106-.5 
Moriori 114-6 112-6 
Burmese A 117-7 111-7 
B 116-7 in-9 
C 116-9 UO-l 
The above table may be of use as giving the corrected values of the auricular 
height. I have added my Burmese for comparison. It will be seen that it makes 
these results still more incompatible with those of Ranke'for the " Altbayerisch" 
120-8, ? 114-2), or Parson's for Hythe 120, ? 116) and Rothwell 120, ? 116). 
7. Description of Material and Comparison of Direct 
Measurements of the Three Series. 
It was obvious in examining the whole series of 142 skulls that they were of 
rather mixed type, although one type predominated. It seemed desirable there- 
fore to class together in separate groups those skulls which qualitative appreci- 
ation led us to suppose of the same type, reserving the right to combine these 
groups afterwards if the quantitative values obtained did not support this judg- 
ment. The procedure was as follows. Seven children's skulls, which could not be 
sexed, nor used to obtain any means, were withdrawn from the rest. The remaining 
skulls were then put into three groups: A and C, with an intermediate group B, 
which combined some of the features of both. Finally, the skulls in the various 
groups were sexed. It was necessary to leave the sexing until the types were 
separated, since some factors which affect the determination of sex, such as weight, 
size, and general boldness of shape, vary considerably from type to type, and some 
young males of a less robust racial type might easily be mistaken for females of 
a heavier and coarser one. For all this classification of material I have to thank 
Professor Pearson. 
We thus had our series divided up into seven sub-series: three male and three 
female sub-series and one composed of children of various a^es. The grouping was 
determined by the general appearance, more particularly of the face. As con- 
tributing factors to the general resemblance of those classed as Group A, we might 
single out as most noteworthy : the nose, broad at the bridge and flattish, with the 
pyriform aperture also wide, and its lower edge very rounded or even forming a 
double ridge with a negroid groove in between; marked sub-nasal prognathism; 
projecting cheek-bones; head of a massive and rugged type, etc. 
* Biometrika, Vol. I. pp. 426-7, Vol. m. pp. 208-9 and Vol. v. pp. 92-3. I have tested eight each of the 
Naqada and Whitechaisel series and have satisfied myself that the extra 2-5 mm?, were added to their 
measurements by the craniophor defect and must therefore be deducted. 
t Ibid. Vol. vm. pp. 208-9. J Ibid. Vol. xi. p. 90. 
